Adverse Health Outcomes Associated with Hypoglycemia Following Bariatric Surgery

Introduction

Hypoglycemia, typically defined as serum glucose concentration below 70 mg/dL or 54 mg/dL, is an increasingly recognized complication of bariatric surgery. Studies using mixed meal tests or oral glucose tolerance tests have shown that up to 90% of patients experience postprandial hypoglycemia following Roux-en-Y gastric bypass (RYGB) or sleeve gastrectomy (SG). ^{1 –3} Moreover, studies that relied on continuous glucose monitoring (CGM) have revealed that hypoglycemia can occur not only after a glucose load but also during a fasting state, with nocturnal hypoglycemia affecting 38%–68% of patients. ^4,5 Additionally, while in some patients hypoglycemia is accompanied by symptoms such as shakiness and sweating, approximately half of patients remain asymptomatic despite having low blood glucose levels. ⁶

The pathophysiology behind the increased occurrence of hypoglycemia postbariatric surgery is not completely understood, but it is hypothesized to be multidimensional. Maintaining glucose levels within a normal range depends on the balance between glucose appearance (the rate at which glucose enters the circulation) and glucose disappearance (the rate at which glucose is removed from the circulation). Mediated by mechanisms such as accelerated gastric emptying and altered gut peptides, bile acids, and gut microbiome, bariatric surgery disrupts the balance between glucose appearance and disappearance: it augments insulin secretion, enhances β-cell sensitivity, decreases insulin clearance, suppresses hepatic gluconeogenesis, increases glucose uptake in the intestine and adipose tissue, and impairs counter-regulation to low glucose levels (e.g. blunted insulin suppression, reduced glucagon responses). ^7,8 Collectively, these changes may contribute to the development of hypoglycemia following bariatric surgery. ⁹

While the high prevalence of postbariatric surgery hypoglycemia has gained increased attention, its clinical implications remain overlooked. Outside the context of bariatric surgery, particularly in patients with type 1 or type 2 diabetes, research repeatedly shows that hypoglycemia has various negative consequences, such as weight gain, ^10,11 deteriorating mental health, ^12,13 disordered eating behaviors, ¹⁴ poorer sleep quality, ^15,16 reduced work productivity, ^17,18 and a diminished quality of life (QoL). ^12,17

Hypoglycemia is associated with hyperphagia, increased energy intake, reduced energy expenditure, and altered adipose storage in response to hypoglycemic stimuli, all of which can theoretically lead to weight gain. ^10,11,19 In addition, when hypoglycemia occurs, patients are often advised to eat in order to elevate their glucose levels, which further contributes to weight gain. In support, several analyses from the Diabetes Control and Complications Trial—a multicenter, randomized, clinical study aimed at assessing the impact of an intensive treatment regimen on early vascular complications in patients with insulin-dependent diabetes—found that patients who experienced hypoglycemia (blood glucose level <70 mg/dL) had significantly higher total and annual weight gain over a 6-year follow-up period. ^10,11

When accompanied by symptoms such as confusion, shivering, blurred vision, drowsiness, and fatigue, hypoglycemia can create persistent fear, uneasiness, and avoidant actions for possible future hypoglycemic episodes, consequently resulting in psychosocial impairment. A large body of cross-sectional studies and a few longitudinal studies have shown that among patients with type 1 or type 2 diabetes, a history of hypoglycemia (mild or severe) is associated with more severe depressive symptoms, ^{12,13,20 –23} worse QoL in both physical and mental domains, ^12,17,23,24 and reduced working capacity. ^17,18,25,26

In the context of bariatric surgery, the consequences of hypoglycemia have been minimally explored, with only a few reports focusing on weight loss and one study examining QoL. The studies on weight loss yielded mixed results: some identified a significant association between hypoglycemia and less weight loss, ^27,28 but others reported null findings. ^29,30 The QoL study, which included 74 patients who have undergone RYGB, did not detect a link between hypoglycemia and QoL, possibly due to insufficient power. ²⁹ To the best of our knowledge, no study has incorporated outcomes of depression or work productivity, signifying another gap in our understanding of the clinical implications of hypoglycemia following bariatric surgery.

The purpose of this study was to examine the longitudinal associations between hypoglycemia and weight loss, depressive symptoms, QoL, and work productivity in a cohort of patients followed for 7 years after bariatric surgery. The results are expected to assist in clinical consultations regarding the negative effects of hypoglycemia and guide the necessary care for patients experiencing hypoglycemia after bariatric surgery.

Methods

Results

The description of sample characteristics is shown in Table 1. Participants (N = 552) were mostly female (82.4%), White (92.4%), had a household income ≥ $50,000 (59.0%), and had undergone RYGB (71.3%). At the time of surgery, the mean age was 47.5 ± 11.1 years and the mean BMI was 46.6 ± 7.5 kg/m².

During the 84-month follow-up, 20%–30% of patients reported experiencing hypoglycemia in the past 3 months. Among those patients, the mean frequency of hypoglycemia in the last 7 days is approximately 2, the mean number of severe hypoglycemic episodes in the past 3 months is less than 1, and more than 95% experienced symptomatic hypoglycemia (Table 2). Linear mixed models (Table 3) indicated that, compared to patients who did not experience hypoglycemia in the past 3 months, those who did had higher depressive symptoms (β = 2.4; 95% CI: 1.7, 3.0), lower scores in physical (β = −4.2; 95% CI: −5.1, −3.3) and mental (β = −3.4; 95% CI: −4.4, −2.4) QoL, and reduced work productivity in the domains of presenteeism (β = 5.8; 95% CI: 3.4, 8.2), work productivity loss (β = 5.6; 95% CI: 2.6, 8.6), and activity impairment (β = 8.8; 95% CI: 6.0, 11.6). In contrast, hypoglycemia occurrence did not appear to influence weight loss (β = −1.4; 95% CI: −2.5, 2.2).

Among participants who reported having hypoglycemia at least once during the 84-month follow-up, a higher frequency of hypoglycemia in the past 7 days was associated with more depressive symptoms (β = 0.7; 95% CI: 0.5, 0.9), worse physical (β = −1.1; 95% CI: −1.4, −0.8), and mental (β = −1.0; 95% CI: −1.3, −0.6) QoL, and reduced work productively, including increased presenteeism (β = 1.7; 95% CI: 0.7, 2.8), greater work productivity loss (β = 1.6; 95% CI: 0.3, 2.9), and more activity impairment (β = 8.8; 95% CI: 6.0, 11.6; Table 4). Similarly, a greater number of severe hypoglycemic episodes in the past 3 months (Table 5) and experiencing symptomatic versus nonsymptomatic hypoglycemia (Table 6) were each associated with more depressive symptoms and poorer physical and mental QoL.

Discussion

This study used data from a multisite cohort study to examine potential negative outcomes of hypoglycemia during a 7-year follow-up period after bariatric surgery. Results suggested that the occurrence of hypoglycemia was associated with more severe depressive symptoms, worse QoL, and reduced work productivity. Furthermore, these negative consequences increased with the increase in the frequency of hypoglycemia, number of severe hypoglycemia episodes, and the presence of symptomatic hypoglycemia.

Approximately 20%–30% of the patients self-reported experiencing hypoglycemia during the past 3 months. The true prevalence of hypoglycemia is difficult to determine due to the different diagnostic methods available. The traditional diagnostic criteria require meeting the Whipple triad (symptoms consistent with hypoglycemia, documented low plasma glucose levels, and symptom resolution after glucose intake). Since the assessment used in LABS-2 did not include laboratory glucose confirmation or symptom resolution upon glucose normalization, the observed prevalence of hypoglycemia is higher than the 5%–10% reported in studies using the Whipple triad. On the contrary, the increasing adoption of CGM in bariatric research has revealed a high prevalence of asymptomatic and nocturnal hypoglycemia. Therefore, because the nature of the question used in this study primarily excludes those who did not experience or who were not aware of hypoglycemic symptoms, our reported prevalence is lower than those reported in CGM-based studies. ^4,5 Further research should consider incorporating both modalities and assessing whether the impact of hypoglycemia on health outcomes differs based on diagnosis via the Whipple triad or CGM.

The findings that link hypoglycemia with depressive symptoms and reduced QoL are consistent with existing research on type 1 or type 2 diabetes, where multiple studies have shown a dose-dependent relationship between hypoglycemia frequency, regardless of its severity, and both depression and QoL. ^{12,21 –23} For instance, a survey of a representative sample of patients with type 2 diabetes in the United States found that respondents who reported hypoglycemia at least once in the past 12 months had significantly higher depression scores and lower QoL scores for both physical and mental health than those without hypoglycemia, after adjusting for covariates such as comorbid conditions. ¹² Furthermore, among those reporting hypoglycemia in the past 4 weeks, mean depression scores increased and mean QoL scores (physical domain) decreased as the number of hypoglycemic episodes increased. ¹²

In the context of bariatric surgery, we are not aware of any studies addressing depression, and only one study that assessed QoL. This single study explored factors influencing QoL 12 years after RYGB and found that hypoglycemia did not emerge as a significant factor in the univariate analysis. ²⁹ However, a major limitation of this study is its small sample size (36 patients assessed for hypoglycemia), which restricts its statistical power to detect potential associations. Additionally, the study focused solely on postprandial hypoglycemia and did not account for episodes occurring in the fasting state, which may also adversely affect QoL.

Our study findings have clinical significance in that although bariatric surgery is known to decrease depression and enhance QoL, a considerable number of patients do not experience these benefits, exhibiting either worsened depression/QoL or a relapse after initial improvement. ^36,37 Commonly cited reasons for this include inadequate weight loss, weight regain, recurrence of obesity-related comorbidities, and surgical side effects such as hair loss and excessive skin. ^36,37 Our analysis suggests that hypoglycemia may significantly impact patient’s well-being and should be considered a target for intervention to improve surgical outcomes.

Participants in our study who experienced hypoglycemia following bariatric surgery reported decreased work productivity, as evidenced by higher presenteeism, work productivity loss, and activity impairment. This observation aligns with findings in the diabetes population. For example, an analysis from the US National Health and Wellness Survey—a cross-sectional, web-based survey of adults aged 18 and older—found that respondents with type 2 diabetes who had recent hypoglycemic episodes in the past 3 months reported significantly higher rates of absenteeism, presenteeism, and overall work productivity impairment compared to those with nonrecent or no hypoglycemia. ¹⁷ Additionally, other studies have noted that following nonsevere hypoglycemic events, 10%–30% of patients with type 1 or type 2 diabetes reported disruptions such as arriving late to work, leaving early, or losing work time. ^18,25,26,38 The link between hypoglycemia and reduced work productivity can be partially explained by impaired cognitive performance. For example, one recent study assessed cognitive function in patients post-RYGB using a series of neurocognitive tests. ³⁹ It found that those with versus without hypoglycemia displayed greater challenges in completing tasks related to semantic processing (e.g. category fluency, category switching, and category switching accuracy), cognitive flexibility, and word retrieval.

No significant associations were found between hypoglycemia and weight loss. This finding adds evidence to the limited and inconsistent literature regarding whether hypoglycemia is a risk factor for reduced weight loss or weight regain following bariatric surgery. In line with our findings, one prospective cohort study (N = 50) reported that hypoglycemia, identified through self-reported symptoms or medical diagnosis, was not linked to weight regain at 12 years after RYGB. ²⁹ Conversely, another study examined postprandial hypoglycemia—screened by hypoglycemic symptoms and confirmed through capillary glucose levels—and weight loss following RYGB or SG, finding that postprandial hypoglycemia was associated with significantly less weight loss at 2 years postsurgery. ²⁷ The mixed findings can be explained by the various methodological differences among studies, including measurement of hypoglycemia, measurement of weight, length of follow-up, and covariates adjusted in analysis. Specifically, while our study had a long-term follow-up with multiple hypoglycemia assessments and predominantly objective weight measures, others relied on a single measurement of hypoglycemia and retrospective recall of weight during a short-to-medium period of observation. ^27,28 Moreover, factors related to eating, such as energy intake, dietary composition, and appetite, are crucial when exploring the relationship between hypoglycemia and weight outcomes. However, none of the studies included these variables, which could also contribute to the observed variability in results.

The main strengths of this study are the relatively large sample size, the longitudinal design with repeated hypoglycemia reports and weight measures, and the use of time-lagged techniques in the statistical models that ensure the temporality of the association. Despite this, several limitations should be noted. First, the question used to determine the occurrence hypoglycemia in our study did not offer instructions for its definition or manifestation, and patients’ responses were not validated through laboratory tests. This could introduce misclassification bias, as patients might have misattributed other disturbances, such as dumping syndrome, to hypoglycemia or failed to recognize that the symptoms they experienced were due to hypoglycemia. To mitigate this bias, we incorporated several other dimensions of hypoglycemia beyond occurrence, including frequency, number of severe episodes, and symptoms, which helped strengthened the validity of study findings. Specifically, patients who frequently experience hypoglycemia or report severe episodes are more likely to have true hypoglycemia than those who report only a single mild episode. Thus, the dose-response association between the frequency of hypoglycemia and severe hypoglycemia with depression and QoL adds confidence that the association between hypoglycemia occurrence and these health outcomes is real. Second, our study considered various potential confounders such as age, gender, and presurgery BMI. However, several other factors that could affect patients’ glycemic levels, mental health, and weight status, such as genetic predispositions, lifestyle behaviors, and concomitant medication use, were not included. Third, patients in the LABS-2 cohort were recruited and assessed over a decade ago, so their characteristics and the types of surgeries they underwent may not reflect the current landscape. Today, SG has replaced RYGB and become the most common type of bariatric surgery. Additionally, patients now tend to have a lower presurgery BMI, as national guidelines have recently lowered the indication for bariatric surgery to all patients with a BMI ≥35 kg/m² or ≥30 kg/m² with accompanying metabolic diseases. ⁴⁰

In conclusion, this secondary analysis of the LABS-2 study found that postbariatric surgery self-reported hypoglycemia was associated with more depressive symptoms, worse QoL, and lower work productivity. Although bariatric surgery generally improves these outcomes, the findings of this study caution that hypoglycemia could undermine the benefits achieved through the procedure, highlighting critical role of glycemic control in preserving overall well-being after bariatric surgery. Future studies employing Whipple triad or CGM to precisely examine hypoglycemia on a broader range of outcomes, such as cognitive function and suicidal ideation, are needed.